Transparency is vital to support public confidence and maintain trust in statistics and those producing them. A lack of transparency can result in confusion about where numbers have come from or can lead to accusations of manipulating the data. Our principles of intelligent transparency are equality of access, understanding and leadership. While there is a greater understanding of the need for transparency, there is more work to do to fully embed the principles of intelligent transparency consistently across government, especially in relation to equality of access. We are still seeing concerns raised with us via our casework process in relation to Ministers and other government officials quoting unpublished figures in the public domain, most commonly when these figures are based on management information.

In June 2022, the Department for Work and Pensions posted a tweet and press release claiming that over half a million people had been helped into work by the Way to Work campaign, meeting its target. The ‘half a million’ figure was taken from an answered Parliamentary Question, using management information, which did not contain a clear explanation of how the Way to Work target was defined, how it would be measured, and the methods used to support claims that the target had been reached. In line with our expectations on the use of management information, we highlighted the need for more formal structured reporting to ensure equality of access, including appropriate explanations of context and sources to support public use such as in this case.

In November 2022, we wrote to the Home Office regarding the transparency of its statistics following concerns raised with us about the use of unpublished data and statistics by Ministers. These concerns were in reference to claims made about those arriving on small boats, the number of asylum caseworkers and the number of adult male migrants claiming to be children. We urged the Home Office to review its ways of working to ensure that the transparent release and use of data and statistics became the default in future.

Where unpublished statistics are quoted in the public domain, this is often reported to be the result of poor communication between the statistical teams and other areas of the government department such as communications or the ministerial private office. Statisticians have a key role to play in proactively supporting and championing intelligent transparency in their departments. We have heard of producers taking a variety of approaches to minimise the risk of unpublished statistics being quoted in the public domain, including embedding analytical teams within private offices, setting out agreements on the use of statistics with departmental communications team and providing training on the use of statistics to the wider government department.

Whilst the production of high quality of statistics is important, the way statistics are communicated to users and the public more widely is key to supporting understanding and appropriate use. As with equality of access, communication of statistics is a common theme in our casework, and we have seen an increase in the number of concerns raised with us throughout the last year. These concerns have been spread across all of our domains, with the majority focused in our Economy, Health and Social Care, and Population and Society domains.

We regularly investigate concerns raised with us through casework and publicly intervene when appropriate. Our intention with any intervention is to deliver positive change, whether that is an improvement in the production or communication of official statistics, or a more responsible use of statistics in public debate. For example, in February 2023, we wrote to HM Treasury about its chart on inflation which we had judged to give a misleading impression of the scales of the deceleration. In this case, HM Treasury responded positively and promptly. Another example of our public interventions is a blog that we published in response to the poor design of a question in the weights and measures consultation from the Department for Business, Energy and Industrial Strategy.

It can be difficult for statisticians to influence how statistics are used once they have been published, with social media a factor in some cases. Statisticians should rise to this challenge by engaging widely with people who use statistics, including media outlets, ensuring a clear understanding of the statistics, to minimise the risk of potential misuse.

Producers primarily present statistics to external users through statistical bulletins. The quality of these publications varies across government departments and there is an increasing view that they are often too lengthy and wordy. The presentation and the usability of publications has a significant impact on the ability of users to easily draw reliable conclusions as well as the risk of misuse. Several statistical bulletins such as the ONS’s Consumer Price Inflation release were highlighted to us as strong examples of clearly drawing out key messages. Producers are exploring alternative ways to communicate information away from the traditional statistical bulletin. For example, the Northern Ireland Department for Communities created the Northern Ireland Local Labour Market Insight dashboard to bring together labour market related indicators from different data sources in an interactive and visual way.

Increasingly government departments and producers are engaging with users via other methods such as podcasts, blogs and social media. For example the Office of Rail and Road produces The Rail and Road Pod and engages with the public through question and answer sessions on Twitter when statistics are released. The ONS also produces a monthly podcast called Statistically Speaking. Many producers have dedicated statistics Twitter accounts such as Welsh Government’s Statistics for Wales Twitter account. While there can be a reluctance from some government departments to provide greater public platforms to statisticians, we consider that these can be a key component to supporting appropriate use of statistics when done in an official capacity.

Another important aspect of communicating statistics is conveying the uncertainty around estimates. Communicating uncertainty is essential in improving the interpretation of the statistics and clarifying what the statistics can and cannot be used for. We have found that the communication of uncertainty is inconsistent across different types of statistical outputs, with statistical bulletins and methodological documents generally better than data tables, data dashboards, downloadable datasets and infographics.

When producers are exploring new ways to communicate statistics, they should ensure that key methodological and contextual information is clearly provided. Research from the Winton Centre and the Economic Statistics Centre of Excellence (ESCoE) found that communicating uncertainty does not reduce trust in the statistical estimates or the producer of the statistics. We recognise that communicating uncertainty is not always an easy task for producers and we plan to produce more guidance and support in addition to the resources and tools already available for producers.

In February 2023, we published research on the factors that are important for effective communication of statistics accompanied by a think piece on statistical literacy. We consider that statistical literacy should not be viewed as a deficit that needs to be fixed, but that producers of statistics should focus on how best to publish and communicate statistics in a way that can be understood by audiences with varying skill levels and abilities. ESCoE’s ONS-funded research programme into the communication and value of economic statistics demonstrates how producers are investing in developing communication methods.